Download Report

Environmental Pollution 203 (2015) 89e96
Contents lists available at ScienceDirect
Environmental Pollution
journal homepage: www.elsevier.com/locate/envpol
Seabirds, gyres and global trends in plastic pollution
Jan A. van Franeker a, *, Kara Lavender Law b
a
b
IMARES, Wageningen-UR, P.O. Box 167, 1790 AD Den Burg (Texel), Netherlands
Sea Education Association, P.O. Box 6, Woods Hole, MA 02543, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 27 November 2014
Received in revised form
20 February 2015
Accepted 25 February 2015
Available online
Fulmars are effective biological indicators of the abundance of ﬂoating plastic marine debris. Long-term
data reveal high plastic abundance in the southern North Sea, gradually decreasing to the north at
increasing distance from population centres, with lowest levels in high-arctic waters. Since the 1980s,
pre-production plastic pellets in North Sea fulmars have decreased by ~75%, while user plastics varied
without a strong overall change. Similar trends were found in net-collected ﬂoating plastic debris in the
North Atlantic subtropical gyre, with a ~75% decrease in plastic pellets and no obvious trend in user
plastic. The decreases in pellets suggest that changes in litter input are rapidly visible in the environment
not only close to presumed sources, but also far from land. Floating plastic debris is rapidly “lost” from
the ocean surface to other as-yet undetermined sinks in the marine environment.
© 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
Keywords:
Fulmarus glacialis
Plastic ingestion
Marine debris
North Atlantic subtropical gyre
Industrial plastic pellets
1. Introduction
Ingestion of marine debris by wildlife, and that of plastics by
seabirds in particular, has been widely documented. Reviews (e.g.
Laist, 1997; Derraik, 2002; Katsanevakis, 2008; Kühn et al., in
press) illustrate the extent of plastic ingestion, but do not evaluate spatial patterns and trends in abundance of marine litter. The
northern fulmar Fulmarus glacialis was among the earliest seabird
species reported to ingest marine plastic debris. Fulmars belong to
the tubenosed bird families of albatrosses and petrels (Procellariiformes). They only come ashore to breed and never forage on
land or in fresh water but exclusively far out to sea. Fulmars have a
wide distribution over the northern North Atlantic and Paciﬁc
Oceans with a population estimated at 15e30 million individuals
(BirdLife International, 2014). Early papers suggested temporal
and spatial differences in accumulated plastics in fulmar stomachs. An abundance of 1e2 particles per fulmar stomach in the
North Sea in the early 1970s (Bourne, 1976) changed to more than
10 plastic particles per stomach by the 1980s (Furness, 1985; Van
Franeker, 1985). Van Franeker (1985) observed an average of 12
plastic particles in fulmars from the North Sea, but less than 5 in
fulmars from the presumably cleaner arctic breeding locations of
Bear Island (74 Ne19 E) and Jan Mayen (71 Ne8 W). Similarly,
* Corresponding author.
E-mail address: [email protected] (J.A. van Franeker).
the difference of only 2.8 plastic particles in fulmars from Alaska
(Day, 1980; Day et al., 1985) compared to 11.3 particles in fulmars
from California (Baltz and Morejohn, 1976) was explained by
higher pollution in waters off the densely populated California
coast. Close relatives of the fulmar living in the Antarctic had still
lower levels of ingested plastics, in which species migrating to
northern areas during winter contained more plastic than the
resident species living in pristine Antarctic waters year round
(Van Franeker and Bell, 1988).
These early studies assumed that plastic abundance in seabird
stomachs reﬂected local or regional pollution levels, which could
then be used to map spatial patterns and to monitor changes over
time in ocean plastic pollution. However, as most datasets were no
more than instantaneous point measurements, there was little
insight into potentially biasing variables affecting quantities of
plastics in bird stomachs. A ﬁrst evaluation of such variables found
that trends over time (1980se2000) in beached fulmars from the
Netherlands were not affected by body condition, sex of the birds,
seasonal variations, or likely breeding region (Van Franeker and
Meijboom, 2002). Only age of birds was found to be a factor in
plastic ingestion, with young and immature birds consistently
having a higher average plastic load in the stomach than adults. For
monitoring purposes, when age composition of samples shows no
structural change towards older or younger birds over time, samples of combined age groups can be used.
Fulmars are now a formal marine litter indicator in OSPAR
(Oslo/Paris Convention for the Protection of the Marine
http://dx.doi.org/10.1016/j.envpol.2015.02.034
0269-7491/© 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
90
J.A. van Franeker, K.L. Law / Environmental Pollution 203 (2015) 89e96
Environment of the North-East Atlantic) and the European MSFD
(Marine Strategy Framework Directive) (OSPAR, 2008, 2010; EC,
2008, 2010; Galgani et al., 2010; MSFD-TSGML, 2013) with results published in peer reviewed literature (Ryan et al., 2009; Van
Franeker et al., 2011). The policy target or ‘Ecological Quality
Objective (EcoQO)’ for an ecologically acceptable level of marine
debris in the North Sea has been deﬁned as fewer than 10% of
beached fulmars in the North Sea having more than 0.1 g of plastic
(OSPAR, 2010). Here we present new information on temporal and
spatial scales in plastic pollution in fulmar stomachs, which will
reﬁne their use as an indicator. Few datasets can conclusively
determine that seabird stomach contents accurately reﬂect environmental abundance of plastic marine debris. In the North Sea,
there are no direct measurements of abundance of plastic debris
in seawater and, although predicted by oceanographic models
(Maximenko et al., 2012; Van Sebille et al., 2012), few data exist to
conﬁrm the lower abundance of ﬂoating plastic debris at high
latitudes (Cozar et al., 2014; Ryan et al., 2014).
While not co-located, one dataset covering an almost similar
time span as the North Sea fulmar study does exist: Sea Education
Association (SEA) has sampled small ﬂoating plastics in the western
North Atlantic Ocean and Caribbean Sea since 1986. In an analysis
of data from 1986 to 2008, Law et al. (2010) found the highest
abundances of plastics in the centre of the North Atlantic subtropical gyre, as predicted by models.
In this paper, we present a comparative analysis of North Sea
fulmar data and SEA data through 2012. The densely populated and
industrialised North Sea area is primarily a source of marine debris,
where winds and currents export ﬂoating debris and prevent local
accumulation (Neumann et al., 2014). In contrast, the North Atlantic
subtropical gyre is distant from major sources, yet accumulates
ﬂoating marine debris.
2. Methods
2.1. Fulmar study
Fulmars used in long-term studies within the North Sea are
birds found dead on beaches. For the Netherlands, data are available from 1979 onwards; other North Sea countries have participated since 2002. From elsewhere, fulmars accidentally killed in
long-line ﬁsheries and stomachs of birds hunted for human consumption have been used. Early Arctic (Van Franeker, 1985) and
Antarctic studies (Van Franeker and Bell, 1988) used birds collected
for the Zoological Museum of Amsterdam.
Standard methods for bird dissections in the monitoring program are described in Van Franeker (2004). Stomach contents are
rinsed in a sieve with a 1 mm mesh and sorted under a binocular
microscope. The 1 mm mesh was selected because smaller particles
are extremely rare in the stomach (Bravo Rebolledo, 2011) and
because smaller meshes clog easily. Plastic items were visually
identiﬁed under binocular microscope and categorized as either
industrial or user plastics. Industrial plastics are often referred to as
pre-production or resin pellets, ‘nurdles’ or ‘mermaids tears’ and
are the raw granular stock from which user objects are made by
melting the granules, with additives giving the plastic its desired
characteristics. User plastics are often fragments of larger objects.
Subcategories of litter are counted and dried at room temperature
for at least 2 days before weighing to an accuracy of 0.0001 g. Data
allow analyses for subcategories of litter or higher groupings by: i)
the percentage of birds having litter in the stomach (incidence or
frequency of occurrence), ii) number of items, or iii) total mass of
litter. Number and mass are always given as population averages,
meaning that all birds, including those with zero debris in the
stomach, are included in the calculation.
Methodological details are provided in Van Franeker et al. (2011)
and the Online Supplement. In the current analysis, time series for
the Netherlands have been updated with results up to 2012 (total
973 birds). For other locations around the North Sea, data in 2012
were not yet available and data are presented up to 2011.
2.2. Gyre study
SEA has sampled small ﬂoating plastics in the western North
Atlantic and Caribbean Sea since 1986. For the current analyses
earlier published data (through 2008 in Law et al., 2010) were
extended through 2012. Samples were collected with neuston nets
and archived by SEA undergraduate students and faculty scientists.
SEA cruises mostly follow annually repeated cruise tracks. The
neuston net has a 1.0 m 0.5 m mouth, a 335-mm mesh, and is
towed at the air-sea interface, in principle sampling half its height
submerged (25 cm). The net is towed off the port side of the vessel
to avoid interference by the ship's wake. Tow duration is typically
30 min at an estimated speed of two knots, giving a nominal tow
length of one nautical mile (1.852 km). However, sampling may
differ by conditions, and actual tow length was measured either
with a taffrail log towed behind the ship or from GPS coordinates.
Plastic particle concentration is computed as total number of pieces
collected, divided by the tow area (tow length*1-m net width), and
reported in units of pieces per km2. The area sampled during a tow
is a small fraction of a square kilometre; when scaled up, the
minimum non-zero concentration recorded is ~540 pieces km2
(one piece in a 1.85 km-long tow). Potential bias from the small
sampling area was tested by comparing averages from individual
tows (with associated standard errors (SE)) to averages derived
from counts of grouped data (total number of items divided by total
area sampled in a year; no SE). Differences were relatively minor, so
here we use values from individual tows. Similar to fulmar data, all
calculations for averages include the net tow observations with
zero plastics. The dataset contained 7165 net tows but observations
east of 50 W (only visited twice; 91 tows), early records that did not
distinguish between industrial and user particles (230 tows), and
likely data entry errors with more than 10 industrial particles but
zero user particles (27 tows) were omitted. The remaining dataset
had 6817 net tows east of 50 W from 1987 to 2012. The analyses in
this paper focus on 2624 records in arbitrarily chosen limits of the
most frequently sampled high density area referred to as the central gyre, between 20 N and 40 N and 60 W to 80 W. Plastic
densities in this centre were about three times higher than those
outside and are expected to more clearly show proportional
abundances and trends over time. The Online Supplement provides
details of backgrounds of data restrictions and tabulates results also
for the unrestricted dataset.
Data graphs for both datasets use 5-year running averages,
each time calculated from all individual birds or net tows within
the period (i.e. not from annual averages). We refrained from
using annual averages because of occasional small samples,
short-term variations and individual outliers. In running average
graphs, the lines connecting data-points are only provided as a
simple visualisation of patterns or trends and have no statistical
meaning.
Temporal trends were evaluated by GAMM (Generalized Additive Mixed Models) using R version 3.0.3 (R Core Team, 2014; Wood,
2011). Where GAMM estimates ‘Effective Degrees of Freedom (edf)
as 1, the correlation may be considered linear (Wood, 2001). Higher
edf indicates more complicated non-linear relationships (Zuur
et al., 2009). Signiﬁcance of all trends was tested by simple linear
regression, ﬁtting log-transformed values of plastic abundance
from individual birds or neuston tows on the year of collection
using Genstat 17th Edition. The test statistic is a t-score for slope
J.A. van Franeker, K.L. Law / Environmental Pollution 203 (2015) 89e96
91
clearance rates of plastics from stomachs and b) investigates the
consistency in small-scale spatial patterns of stomach contents.
3.2. Retention time of plastics in stomachs
Fig. 1. Change in abundance of plastic debris in stomach contents of cape petrels after
return from northern wintering areas to their Antarctic breeding area in Wilkes Land
(66 Se110 E) from late October onwards.
and standard error of the slope estimated by the regression. For
evaluation of regional differences, plastic data were ﬁtted in a
negative binomial generalized linear model with region included as
a factor, and the test statistic is a t-score based on residual variance
for the region (Genstat 17th Edition).
3. Results
3.1. Fulmar study
In order for stomach contents to reﬂect location-speciﬁc pollution levels, birds must forage in a certain area for time periods long
enough to integrate debris encounters, and plastics must disappear
from the stomach quickly enough to ensure that amounts of debris
regain a new local balance when the birds migrate to another area.
Lacking straightforward data on those issues, an indirect approach
is used that a) evaluates information on the residence time and
Unlike gulls, fulmarine petrels do not usually regurgitate indigestible hard items. They only spit out stomach contents in fear, in
ﬁghts, or when feeding their chicks. When they do spit, only materials from the glandular ﬁrst stomach (proventriculus) are lost as
the narrow passage to the second muscular stomach (gizzard)
prevents materials in the gizzard from returning to the proventriculus (Ryan and Jackson, 1986). Most plastic particles accumulate
in the muscular gizzard, where all hard food or debris items are
ground up until they wear down or fragment into sizes small
enough to pass into the intestines. In a study of several species of
Antarctic fulmarine petrels, Van Franeker and Bell (1988) evaluated
changes in stomach contents throughout the breeding season. The
non-polluted character of the local Antarctic area was demonstrated by the fact that the non-migratory species had virtually no
plastic in the stomach in any time of the breeding season, whereas
species migrating north in winter, such as the cape petrel Daption
capense, returned with considerable amounts of ingested plastic. In
their Antarctic breeding area between December and January, cape
petrels lost 80e90% of plastics from their stomachs in just over one
month. However, cape petrels start to arrive in Antarctica in late
October: one late October bird from the study area had a stomach
content similar to that of cape petrels collected in winter off South
Africa (Ryan, 1987) and the Crozet Islands (Van Franeker and J.-K.
Jensen, unpublished). Plastic abundance in these ‘pre-breeding’
cape petrels, compared to birds collected in the Antarctic breeding
location in December, indicates that during this initial ~1.5 months
of local foraging the number and mass of plastic items decreased by
80%e90% (Fig. 1; details in Online Supplement). We conclude that
the rapid losses of plastics were the result of size reductions in the
birds' gizzards and eventual excretion, and that little or no plastic
was ingested while foraging in the Antarctic. A similar rapid
reduction was observed for squid beaks in the stomachs of all
species of fulmarine petrels in the study. Squid beaks are made of
chitin, a natural equivalent of synthetic polymers, and of similar
resistance. Squid are prevalent in winter foraging grounds but are
Fig. 2. Latitudinal patterns in fulmar EcoQO performance (proportion of fulmars having >0.1 g plastic in the stomach) in North Atlantic and Paciﬁc Oceans. (a) Bond et al. (2014), (b)
this study, (c) Kühn and Van Franeker (2012), (d) combined from Mallory et al. (2006), Mallory (2008) and Provencher et al. (2009) with additional information from the authors, (e)
Nevins et al. (2011), (f) Avery-Gomm et al. (2012). Details in Online Supplement.
92
J.A. van Franeker, K.L. Law / Environmental Pollution 203 (2015) 89e96
highest plastic abundance in fulmars from the Channel and lower
average plastic abundances in more northern sub-regions and the
Faroe Islands, with increasing distance from heavily industrialized
and populated areas.
3.4. Time trends 1980s to 2012 in fulmar plastic ingestion
Fig. 3. Regional trends in fulmar EcoQO performance (proportion of fulmars
having > 0.1 g plastic in the stomach) over time in North Sea regions and the Faroe
Islands (Updated from Van Franeker and the SNS Fulmar Study Group (2013); details in
Online Supplement).
rare near the breeding colonies (cf. Jarman et al., 2013). Squid beaks
disappeared at an average rate of 72% between December and
January (details in Online Supplement), consistent with observations of ingested plastic.
3.3. Consistency in regional patterns
For northern fulmars, large-scale spatial patterns in stomach
contents observed in the 1980s have been conﬁrmed by recent
studies that used the EcoQO methods. In the North Sea, on average
about 60% of fulmars exceed the critical EcoQO level of 0.1 g of
plastic. Further north, around the Faroe Islands, this ﬁgure decreases
to about 40%. Incidental data from locations at greater distance from
populated industrialized areas support continuation of this pattern
in both the North Atlantic and in the North Paciﬁc (Fig. 2).
Data for EcoQO performance within the North Sea reveal relatively consistent spatial patterns at even smaller subregional scales
(Fig. 3). The 5-year running averages for EcoQO performance do
suggest small changes over time such as increases in the Channel
area and northeast England, but none of the trends is signiﬁcant.
Ingested plastic mass in fulmars from the Channel area differed
signiﬁcantly from that in the Faroes (p < 0.001) and SE-North Sea
(p ¼ 0.026), but not from the other regions, likely due to lower
sample sizes. The major point illustrated is that differences between sub-regions within the North Sea are fairly consistent, with
Plastic abundance in fulmar stomachs from the Netherlands has
shown strong but erratic changes from the 1980s onwards. In the
standard EcoQO approach, plastic abundance is evaluated in terms
of mass because mass is considered to be more ecologically relevant
than numerical abundance (Van Franeker et al., 2011). Numerical
and mass trends do not always match because particles of user
plastics in fulmar stomachs have become smaller over time (Online
Supplement). The data (1979e2012, n ¼ 973) suggest an increase in
ingested plastic from the mid-1980s to peak values in the mid1990s in both mass and number, followed by a decrease in mass
towards the turn of the century, but not in number. Finally, over the
past decade, number and mass of plastics are apparently stable.
These non-linear patterns in total plastic abundance (industrial
plus user plastics) are visible in 5-year running averages (Fig. 4) but
GAMM analysis only supports non-linear change in number of
particles (edf ¼ 1.7, p ¼ 0.06) and not in mass (edf ¼ 1, p ¼ 0.07;
Online Supplement). Linear regression of total plastics over the
entire time series suggests a strong and signiﬁcant numerical increase (p < 0.001, Fig. 4A), but a weakly signiﬁcant decrease in mass
(p ¼ 0.03, Fig. 4B). Remarkable differences exist between industrial
and user plastics. User plastics dominate the overall pattern (Fig. 5A
and B) and follow non-linear changes described by GAMM for both
number of particles (edf ¼ 2.3, p ¼ 0.005) and mass (edf ¼ 2.9,
p ¼ 0.009). However, GAMM analyses indicate that temporal trends
in industrial plastic (Fig. 5A and B) should be considered linear
(number of particles edf ¼ 1, p ¼ 0.07; mass edf ¼ 1, p ¼ 0.15). Linear
regression indicates a highly signiﬁcant decrease of industrial
plastics (p < 0.001 for both mass and number). This decrease represents an almost 75% reduction in average number of industrial
plastics in stomachs of fulmars found in the Netherlands (from ±8
industrial plastics per stomach in the ﬁrst half of the 1980s to less
than 3 in the 2000s).
3.5. Time trends 1987e2012 in plastic abundance in the North
Atlantic subtropical gyre
In the central part of the gyre, total plastic abundance by
number of particles (Fig. 6) followed a complex non-linear pattern
Fig. 4. Changes in A. numerical abundance and B. mass of plastics in fulmars from the Netherlands since the 1980s. Data show arithmetic averages ±standard error (SE) by running
5-year averages (i.e. data points shift one year ahead at a time; sample size for 5 year periods is 21 during the 1980s and 204 from the 1990s onward. Data in the early 1990s
were omitted because sample sizes were 10 birds. Details in Online Supplement.
J.A. van Franeker, K.L. Law / Environmental Pollution 203 (2015) 89e96
93
Fig. 5. Dissimilar trends in A. numerical abundance and B. mass of industrial and user plastics in fulmars from the Netherlands since the 1980s. Data show arithmetic averages ±SE
by running 5-year averages. For industrial plastics, high arithmetic averages and large standard errors in the last three pentads were caused by 2 excessive outliers in 2010 and 2011.
Additional log transformation in tests for trends reduces the impact of these outliers, see Online Supplement.
over the period 1987e2012 (n ¼ 2624) with strong variability and
potential peak values in the early 2000s (GAMM: edf ¼ 6.2,
p < 0.001; Online Supplement). However, as in contents of fulmar
stomachs, this pattern is composed (Fig. 7) of a dominant trend in
the abundant user plastics with similar non-linear complexity (edf
6.1, p < 0.001) but a linear correlation indicated for the number of
industrial plastics (edf ¼ 1, p < 0.001). By linear regression, the total
number of particles and the number of user plastics have shown no
signiﬁcant change, but industrial plastics have decreased at a highly
signiﬁcant rate (p < 0.001).
Abundance data as number per km2 obscures the fact that even
in the centre of the gyre an average of only 1.05 industrial plastics
and 18.3 user plastics per tow were observed over the entire time
record, illustrating the large number of tows with zero plastics. As a
consequence, trends can be more strongly visualised in the frequency of occurrence of particles in individual tows (Fig. 8). Within
the central gyre, the percentage of net tows that contained one or
more industrial plastics dropped from ~50% in the 1980s to 10e20%
in recent survey years, a highly signiﬁcant decrease over 25 years of
data (p < 0.001). Overall, the density of industrial plastics in the
central gyre has decreased by about 75% from roughly 1000 to
around 250 particles per km2. User plastics are found in about 80%
of net tows without signiﬁcant changes over time.
Early papers estimated retention times for plastics in seabird
stomachs from 2 to 3 months for ‘soft’ objects to 10e15 months for
“hard” particles (Day, 1980). From a wider range of observations,
Day et al. (1985) concluded that it took an average of 6 months or
more for plastic particles to disappear through wear in the gizzard,
with great variation in rates depending on the number, size, and
type of particles. An even longer retention time for plastic pellets
was inferred by Ryan and Jackson (1987) from experimental work
on chicks of white-chinned petrels (Procellaria aequinoctialis); they
estimated a half-life of at least one year for plastics in the stomachs
of these chicks.
Our data on cape petrels demonstrate that these are serious
overestimates of residence time of plastics in stomachs of petrels,
as supported by studies of seabirds in the Canadian Arctic after
their return from winter ranges. Northern fulmars collected at
Nunavut in the high Arctic (n ¼ 102; data derived from Mallory,
2008) showed an overall 90% decrease in the average number of
plastic particles in the stomach over summer from 8.6 particles/
bird in May, to 3.2 in June, 1.2 in July, and 0.8 in August. The June
and July data represent monthly reductions of more than 60%, a
similar order of magnitude to our ﬁndings. The lowered reduction
Fig. 6. Numerical abundance of plastic particles (n/km2) in the central area of the
North Atlantic subtropical gyre (20e40 N, 60e80 W) from 1987 to 2012 in 5-year
running averages ±SE (minimum sample size per 5 years is 237 tows; total tows 2624).
Fig. 7. Numerical abundances of industrial and user plastics (n/km2) in the central area
of the Atlantic gyre (20-40 N, 60e80 W) from 1987 to 2012 in 5 year running averages ±SE (minimum sample size per 5 years ¼ 237 tows; total tows ¼ 2624).
4. Discussion
94
J.A. van Franeker, K.L. Law / Environmental Pollution 203 (2015) 89e96
Fig. 9. Comparative trends in numerical abundance of industrial plastics in stomachs
of North Sea fulmars and surface densities in the North Atlantic subtropical gyre by
running geometric means over 5-year periods.
Fig. 8. Annual incidences of industrial and user plastics in SEA net tow data from the
central gyre (20-40 N, 60e80 W) from 1987 to 2012 (n ¼ 2624).
rate towards the end of summer (JulyeAugust: 33%) may reﬂect
highly wear-resistant plastics remaining in the stomach or low
rates of local ingestion. In a different species, 13% of thick-billed
murres (Uria lomvia) arrived in their high Arctic breeding colony
with plastic in the stomach, while 2 months later no bird at the
same location had any plastic (Provencher et al., 2010).
The studies of birds moving from polluted wintering areas to
clean(er) foraging zones justify the conclusion that for species
comparable in size and morphology to fulmars, the loss rate of
plastics from their stomachs may be conservatively estimated to be
on the order of 75% per month for harder types of plastic. It is
reasonable to assume that softer sheet-like and foamed plastics
disappear at faster rates. Consequently, it is likely that fulmars can
accumulate or lose e quantities characteristic of local pollution
levels within a few weeks, with faster changes possible for softer
materials.
Fulmars cover distances of around 30 km in an hour, up to a
maximum of 70 km (Falk and Møller, 1995; Weimerskirch et al.,
2001; Mallory et al., 2008; Edwards et al., 2013). Theoretically,
such ﬂight speeds enable birds to cover much of the North Sea in a
few days. However, continuous fast movements are energetically
expensive, and in practise seabirds tend to stay for longer periods
once in chosen foraging areas. From tracking studies during the
breeding season, foraging ranges of breeding fulmars have been
estimated at only 47.5 ± 17.7 (sd) km away from the colony (Thaxter
et al., 2012), in spite of the fact that the maximum observed distance of a breeding bird away from the colony was around 2400 km
during a 15-day journey in the early egg phase (Edwards et al.,
2013). Winter foraging patterns are less well known. Tracking
data show wide dispersal potential, but also indicate fairly limited
daily travel distances. Mallory et al. (2008) recorded an average
travel distance of 84 km/day for high-Arctic Canadian fulmars, but
this included the fast initial southward migration and thus strongly
overestimates movement in the winter foraging zone. Individual
tracks of Paciﬁc fulmars (Hatch et al., 2010) showed considerable
variability in wintering patterns, but quite a few birds showed
behaviour of staying relatively sedentary once in chosen locations,
sometimes returning to the same small area in subsequent winters.
Fulmar tracking data indicating relative short daily movements are
consistent with our ﬁndings on spatial gradients in plastics
abundance (Fig. 3) and stomach residence time of plastics. On
average, stomach contents of fulmars reﬂect the local conditions to
which they adapt on time scales of a few weeks or possibly even
days.
Our analyses indicate similarity in long-term trends in plastic
abundance ingested by a bio-indicator in the North Sea, one of the
source areas for plastic debris in the North Atlantic, and surface
densities in the North Atlantic subtropical gyre, a long-term accumulation area (Law et al., 2010; Maximenko et al., 2012; Van Sebille
et al., 2012). Although Moret-Ferguson et al. (2010) examined
plastic mass in a subset of the gyre data, we have insufﬁcient information to compare fulmar and gyre data using plastic mass, and
thus focus on numerical abundances.
Industrial plastics show highly signiﬁcant decreases throughout
the period of observation, strongest in initial years but continuing
into an overall reduction of about 75% in both datasets over two to
three decades (Fig. 9). These data are consistent with the ‘spot’
observations on abundance of industrial plastics in seabird stomach
contents in other areas. In the western Atlantic, Moser and Lee
(1992) reported half of the plastic items in fulmar stomachs as industrial during the early 1980s, whereas Bond et al. (2014) in recent
years classiﬁed only 6% as industrial plastics. In the North Paciﬁc,
industrial plastics in stomachs of short-tailed shearwaters (Pufﬁnus
tenuirostris) nearly halved from the 1970s to the period 1997e2001
(Vlietstra and Parga, 2002). In the South Atlantic and Indian Oceans,
Ryan (2008) reported 44%e79% decreases in the abundance of industrial plastic particles in 5 tubenosed seabird species from the
1980s to 1999e2006. Thus, there is convincing evidence for a
Fig. 10. Comparative trends in numerical abundance of user plastics in stomachs of
North Sea fulmars and surface densities in the North Atlantic subtropical gyre by
running geometric means over 5-year periods.
J.A. van Franeker, K.L. Law / Environmental Pollution 203 (2015) 89e96
95
land (4.8 million tons; Jambeck et al., 2015). Our time series for
industrial plastics provide ﬁrm evidence that such a mismatch has a
realistic basis and is not due to potential errors in measurements or
models. We do not know to what extent losses from the ocean
surface represent export to other oceanic compartments or to land,
reductions in size to below our level of observation, or possibly true
degradation. Ingestion and stomach processing by wildlife may
well play a role in size reductions and displacement. The hypothesis
that a reduced (but continuing) rate of input of plastics leads to
reduced numbers of particles in marine surface waters does not
mean that current input levels do not cause harm to food-chains or
ecosystems. The critical question, ‘Where is all the plastic?’
(Thompson et al., 2004), including the uncertainty on impacts, remains unanswered. However, our observations do suggest that a
reduction in the input of plastic debris to the sea is an observable
and effective way to at least begin solving this pollution problem.
Fig. 11. Numerical proportion of industrial plastics among all plastic particles in the
central Atlantic gyre and in northern fulmars of the North Sea by running 5-year periods (arithmetic average number of industrial plastics divided by arithmetic average
total number of plastic particles).
strongly reduced abundance of industrial plastics in the surface of
the global oceans from the 1980s into the 21st century. Our North
Sea fulmar data show that input of industrial plastics in one of the
major areas of global plastic production (PlasticsEurope, 2013) has
been reduced.
User plastics, on the other hand, have shown a complex pattern
of increases and decreases in numerical abundance in both fulmars
and the gyre. Fulmars showed an initial strong numerical increase
and subsequent stability, whereas abundance in the gyre ﬂuctuated
without an evident long-term trend (Fig. 10).
The different patterns for industrial and user plastics have led to
a considerable change in composition of plastic in both the gyre and
in fulmars. During the ﬁrst half of the 1980s fulmars had about
equal numbers of industrial and user plastic particles in their
stomachs; currently, user plastics outnumber industrial plastics by
a factor of 10. In the gyre, initially about one in seven particles was
an industrial pellet, but recently only one in about 50. Fig. 11 suggests that the major changes in these ratios occurred before the
turn of the century, and that recently proportions remain fairly
stable.
A tentative explanation for the decrease in industrial plastics
might be found in a response to publicity in the 1970s and 1980s
revealing a global oceanic presence of virgin industrial pellets
(Colton et al., 1974; Wong et al., 1974; Gregory, 1978; Shiber, 1979,
1982; Morris, 1980) and their ingestion by a wide range of marine
wildlife (e.g. Bourne and Imber, 1982; Connors and Smith, 1982;
Day et al., 1985). For the 1980s, no information exists on dedicated measures by industry or transport sectors, but in 1991 the
dedicated Operation Clean Sweep campaign was started (U.S. EPA,
1993). The similarity in results of fulmar and gyre data, and published information on seabirds elsewhere, suggests that the
observed trends are embedded in a wider and more general
reduction in the input of industrial pre-production pellets to the
marine environment.
Our analysis shows that reduced input of marine debris in
source areas has observable effects even in accumulation areas far
offshore within a limited number of years. This implies that plastics
disappear from the sea surface on relatively short time scales.
Recent publications (Cozar et al., 2014; Eriksen et al., 2014) reported
lower than expected accumulation of plastic debris in all 5 global
subtropical gyres. Eriksen et al. (2014) estimated that around
270.000 tons of micro- and macro plastic debris ﬂoats in the global
oceans. That quantity represents only about 5% of the minimum of
the estimated annual input of plastic waste into the oceans from
Acknowledgements
This paper had its origin in the Marine Debris working group
convened by the National Center for Ecological Analysis and Synthesis (NCEAS), University of California, Santa Barbara, with support from Ocean Conservancy. We are especially grateful to Steve
Gaines, George Leonard and Nicholas Mallos.
Fulmar studies were supported by the Netherlands Ministry of
Infrastructure and the Environment, EU Interreg, NYK Group
Europe and Ltd Chevron Upstream Europe. Mark Mallory and Jennifer Provencher provided unpublished details of Canadian data.
Jens-Kjeld Jensen provided samples of cape petrels from Crozet. The
Save the North Sea (SNS) Fulmar Study Group consists of too many
people to be named individually, but without whom a project such
as this is impossible. Fokje Schaafsma and Santiago Alvarez Fernandez provided valuable help with GAMM analyses. We thank the
thousands of SEA students and marine and scientiﬁc staff on board
the R/V Westward and the SSV Corwith Cramer. KLL was supported
by the National Science Foundation (OCE-1260403) and Sea Education Association.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.envpol.2015.02.034.
References
Avery-Gomm, S., O'Hara, P.D., Kleine, L., Bowes, V., Wilson, L.K., Barry, K.L., 2012.
Northern fulmars as biological monitors of trends of plastic pollution in the
eastern North Paciﬁc. Mar. Pollut. Bull. 64, 1776e1781.
Baltz, D.M., Morejohn, G.V., 1976. Evidence from seabirds of plastic pollution off
central California. West. Birds 7, 111e112.
BirdLife International. Species Factsheet: Fulmarus glacialis. Downloaded from.
http://www.birdlife.org. on 04/05/2014
Bond, A.L., Provencher, J.F., Daoust, P.-Y., Lucas, Z.N., 2014. Plastic ingestion by fulmars and shearwaters at Sable Island, Nova Scotia, Canada. Mar. Pollut. Bull. 87,
68e75.
Bourne, W.R.P., 1976. Seabirds and pollution. In: Johnston, R. (Ed.), Marine Pollution.
Academic Press, London, pp. 403e502.
Bourne, W.R.P., Imber, M.J., 1982. Plastic pellets collected by a Prion on Gough Island, Central South Atlantic Ocean. Mar. Pollut. Bull. 13, 20e21.
Bravo Rebolledo, E., 2011. Threshold Levels and Size Dependent Passage of Plastic
Litter in Stomachs of Fulmars (MSc thesis). Wageningen University. Aquatic
Ecology and Water Quality Management group Report no. 008/2011.
Colton Jr., J.B., Knapp, F.D., Burns, B.R., 1974. Plastic particles in surface waters of the
northwestern Atlantic. Science 185 (4150), 491e497.
Connors, P.G., Smith, K.G., 1982. Oceanic plastic particle pollution: suspected effect
on fat deposition in red phalaropes. Mar. Pollut. Bull. 13, 18e20.
zar, A., Echevaria, F., Gonza
lez-Gordillo, J.I., Irigoien, X., Úbeda, B., Hern
Co
andez n, S., Palma, A.T., Navarro, S., García-de-Lomas, J., Ruis, A., Ferna
ndez-deLeo
Puelles, M.L., Duarte, C.M., 2014. Plastic debris in the open ocean. PNAS 111,
10239e10244.
Day, R.H., 1980. The Occurrence and Characteristics of Plastic Pollution in Alaska's
96
J.A. van Franeker, K.L. Law / Environmental Pollution 203 (2015) 89e96
Marine Birds (M.S. thesis). Univ. Alaska, Fairbanks.
Day, R.H., Wehle, D.H.S., Coleman, F.C., 1985. Ingestion of plastic pollutants by
marine birds. In: Shomura, R.S., Yoshida, H.O. (Eds.), Proceedings of the Workshop on the Fate and Impact of Marine Debris, 26-29 November 1984. U.S. Dep.
Commerce, NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-SWFC-54, Honolulu,
Hawaii, pp. 344e386.
Derraik, J.G.B., 2002. The pollution of the marine environment by plastic debris: a
review. Mar. Pollut. Bull. 44, 842e852.
EC, 2008. Directive 2008/56/EC of the European Parliament and of the Council of 17
June 2008 establishing a framework for community action in the ﬁeld of marine
environmental policy (Marine Strategy Framework Directive). Ofﬁcial J. Eur.
Union L 164, 19e40 (25 Jun 2008).
EC, 2010. Commission Decision of 1 September 2010 on criteria and methodological
standards on Good Environmental Status of marine waters (notiﬁed under
document C(2010) 5956) (Text with EEA Relevance) (2010/477/EU). Ofﬁcial J.
Eur. Union L232, 14e24.
Edwards, E.W.J., Quinn, L.R., Wakeﬁeld, E.D., Miller, P.I., Thompson, P.M., 2013.
Tracking a northern fulmar from a Scottish nesting site to the Charlie Gibbs
Fracture Zone: evidence of linkage between coastal breeding seabirds and MidAtlantic Ridge feeding sites. Deep Sea Res. Par II: Top. Stud. Oceanogr. 98,
438e444.
Eriksen, M., Lebreton, L.C.M., Carson, H.S., Thiel, M., Moore, C.J., Borrero, J.C.,
Galgani, F., Ryan, P.G., Reisser, J., 2014. Plastic pollution in the world's oceans:
more than 5 trillion plastic pieces weighing over 250,000 tons aﬂoat at sea.
PLoS ONE 9 (12), e111913.
Falk, K., Møller, S., 1995. Satellite tracking of high-arctic Northern Fulmars. Polar
Biol. 15, 495e502.
Furness, R.W., 1985. Plastic particle pollution: accumulation by Procellariiform
seabirds at Scottish colonies. Mar. Pollut. Bull. 16, 103e106.
Galgani, F., Fleet, D., van Franeker, J., Katsanevakis, S., Mouat, J., Oosterbaan, L.,
Poitou, I., Hanke, G., Thompson, R., Amato, E., Birkun, A., Janssen, C., 2010.
Properties and Quantities of Marine Litter Do Not Cause Harm to the Coastal
and Marine Environment. Report on the Identiﬁcation of Descriptors for the
Good Environmental Status of European Seas Regarding Marine Litter under the
Marine Strategy Framework Directive. MSFD GES Task Group 10, Final Report
19/04/2010.
Gregory, M.R., 1978. Accumulation and distribution of virgin plastic granules on
New Zealand beaches. N. Z. J. Mar. Freshw. Res. 12, 399e414.
Hatch, S.A., Gill, V.A., Mulcahy, D.M., 2010. Individual and colony-speciﬁc wintering
areas of Paciﬁc northern fulmars (Fulmarus glacialis). Can. J. Fish. Aquatic Sci. 67,
386e400.
Jambeck, J.R., Geyer, R., Wilcox, C., Siegler, T.R., Perryman, M., Andrady, A.,
Narayan, R., Law, K.L., 2015. Plastic waste inputs from land into the ocean.
Science 347 (6223), 768e771.
Jarman, S.N., McInnes, J.C., Faux, C., Polanowski, A.M., Marthick, J., Deagle, B.E.,
lie penguin population diet monitoring
Southwell, C., Emmerson, L., 2013. Ade
by analysis of food DNA in scats. PLoS ONE 8 (12), e82227.
Katsanevakis, S., 2008. Marine debris, a growing problem: sources, distribution,
composition and impacts. In: Hofer, T.N. (Ed.), Marine Pollution: New Research.
Nova Science Publishers, Inc., pp. 53e100.
Kühn, S., Van Franeker, J.A., 2012. Plastic ingestion by the Northern Fulmar (Fulmarus glacialis) in Iceland. Mar. Pollut. Bull. 64, 1252e1254.
Kühn, S., Bravo Rebolledo E.L., Van Franeker, J.A. Deleterious effects of litter on
marine life. In: Bergmann, M., Gutow, L., Klages, M. (eds). Marine Anthropogenic
Litter, (in press), Springer Verlag, Berlin.
Laist, D.W., 1997. Impacts of marine debris: entanglement of marine life in marine
debris including a comprehensive list of species with entanglement and
ingestion records. In: Coe, J.M., Rogers, D.B. (Eds.), Marine debris Sources, Impacts and Solutions, Springer Series on Environmental Management. Springer
Verlag, New York, pp. 99e140.
Law, K.L., Moret-Ferguson, S., Maximenko, N.A., Proskurowski, G., Peacock, E.,
Hafner, J., Reddy, C.M., 2010. Plastic Accumulation in the North Atlantic subtropical gyre. Science 329, 1185e1188.
Mallory, M.L., 2008. Marine plastic debris in northern fulmars from the Canadian
High Arctic. Mar. Pollut. Bull. 56, 1486e1512.
Mallory, M.L., Roberston, G.J., Moenting, A., 2006. Marine plastic debris in northern
fulmars from Davis Strait, Nunavut, Canada. Mar. Pollut. Bull. 52, 813e815.
Mallory, M.L., Akearok, J.A., Edwards, D.B., O'Donovan, K., Gilbert, C.D., 2008.
Autumn migration and wintering of northern fulmars (Fulmarus glacialis) from
the Canadian high Arctic. Polar Biol. 31, 745e750.
Maximenko, N., Hafner, J., Niiler, P., 2012. Pathways of marine debris derived from
trajectories of Lagrangian drifters. Mar. Pollut. Bull. 65, 51e62.
Moret-Ferguson, S., Law, K.L., Proskurowski, G., Murphy, E.K., Peacock, E.E.,
Reddy, C.M., 2010. The size, mass, and composition of plastic debris in the
western North Atlantic Ocean. Mar. Pollut. Bull. 60, 1873e1878.
Morris, R.J., 1980. Plastic debris in the surface waters of the South Atlantic. Mar.
Pollut. Bull. 11, 164e166.
Moser, M.L., Lee, D.S., 1992. A fourteen-year survey of plastic ingestion by western
North Atlantic seabirds. Colon. Waterbirds 15, 83e94.
MSFD-TSGML, 2013. Guidance on Monitoring of Marine Litter in European Seas e a
Guidance Document within the Common Implementation Strategy for the
Marine Strategy Framework Directive. EUR-26113 EN JRC Scientiﬁc and Policy
Reports JRC83985, p. 128. http://dx.doi.org/10.2788/99475.
Neumann, D., Callies, U., Matthies, M., 2014. Marine litter ensemble transport
simulations in the southern North Sea. Mar. Pollut. Bull. 86, 219e228.
Nevins, H., Donnelly, E., Hester, M., Hyrenbach, D., 2011. Evidence for increasing
plastic ingestion in northern fulmars (Fulmarus glacialis rodgersii) in the Paciﬁc.
In: Fifth International Marine Debris Conference, Honolulu Hawaii 20e25 Mar
2011. Oral Presentation Extended Abstracts 4.b.3, pp. 140e144.
OSPAR, 2008. Background Document for the EcoQO on Plastic Particles in Stomachs
of Seabirds. OSPAR Commission, Biodiversity Series Publication Number: 355/
2008. OSPAR, London.
OSPAR, 2010. The OSPAR System of Ecological Quality Objectives for the North Sea:
a Contribution to OSPAR's Quality Status Report 2010. OSPAR Publication 404/
2009. OSPAR Commission London, en Rijkswaterstaat VenW, Rijswijk, p. 16
(Update 2010).
Provencher, J.F., Gaston, A.J., Mallory, M.L., 2009. Evidence for increased ingestion of
plastics by northern fulmars (Fulmarus glacialis) in the Canadian Arctic. Mar.
Pollut. Bull. 58, 1092e1095.
Provencher, J.F., Gaston, A.J., Mallory, M.L., O'Hara, P.D., Gilchrist, H.G., 2010.
Ingested plastic in a diving seabird, the thick-billed murre (Uria lomvia), in the
eastern Canadian Arctic. Mar. Pollut. Bull. 60, 1406e1411.
R Core Team, 2014. R: a Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.
org/.
Ryan, P.G., 1987. The incidence and characteristics of plastic particles ingested by
seabirds. Mar. Environ. Res. 23, 175e206.
Ryan, P.G., 2008. Seabirds indicate changes in the composition of plastic litter in the
Atlantic and south-western Indian Oceans. Mar. Pollut. Bull. 56, 1406e1409.
Ryan, P.G., Jackson, S., 1986. Stomach pumping: is killing seabirds necessary? Auk
103, 427e428.
Ryan, P.G., Jackson, S., 1987. The lifespan of ingested plastic particles in seabirds and
their effect on digestive efﬁciency. Mar. Pollut. Bull. 18, 217e219.
Ryan, P.G., Moore, C.J., Van Franeker, J.A., Moloney, C.L., 2009. Monitoring the
abundance of plastic debris in the marine environment. Philosophical Trans. R.
Soc. B 364, 1999e2012.
Ryan, P.G., Musker, S., Rink, A., 2014. Low densities of drifting litter in the African
sector of the Southern Ocean. Mar. Pollut. Bull. 89, 16e19.
Shiber, J.G., 1979. Plastic pellets on the coast of Lebanon. Mar. Pollut. Bull. 10, 28.
Shiber, J.G., 1982. Plastic pellets on Spain's ‘Costa del Sol’ beaches. Mar. Pollut. Bull.
13, 409e412.
Thaxter, C.B., Lascelles, B., Sugar, K., Cook, A.S.C.P., Roos, S., Bolton, M.,
Langston, R.H.W., Burton, N.H.K., 2012. Seabird foraging ranges as a preliminary
tool for identifying candidate marine protected areas. Biol. Conserv. 156 (SI),
53e61.
Thompson, R.C., Olsen, Y., Mitchell, R.P., Davis, A., Rowland, S.J., John, A.W.G.,
McGonigle, D., Russell, A.E., 2004. Lost at sea: where is all the plastic? Science
304 (5672), 838.
U.S. EPA, 1993. Plastic Pellets in the Aquatic Environment - Sources and Recommendations. A Summary. United States Environmental Protection Agency EPA,
842-S-93-001. http://water.epa.gov/type/oceb/marinedebris/upload/2009_11_
23_oceans_debris_plasticpellets_plastic_pellets_summary.pdf.
Van Franeker, J.A., 1985. Plastic ingestion in the North Atlantic fulmar. Mar. Pollut.
Bull. 16, 367e369.
Van Franeker, J.A., 2004. Save the North Sea - Fulmar Study Manual 1: Collection
and Dissection Procedures. Alterra Rapport 672. Alterra, Wageningen.
Van Franeker, J.A., Bell, P.J., 1988. Plastic ingestion by petrels breeding in Antarctica.
Mar. Pollut. Bull. 19, 672e674.
Van Franeker, J.A., Meijboom, A., 2002. Litter NSV e Marine Litter Monitoring by
Northern Fulmars: a Pilot Study. ALTERRA-Rapport 401. Alterra, Wageningen.
Van Franeker, J.A., the SNS Fulmar Study Group, 2013. Fulmar Litter EcoQO Monitoring along Dutch and North Sea Coasts e Update 2010 and 2011. IMARES
Report C076/13. IMARES, Texel.
Van Franeker, J.A., Blaize, C., Danielsen, J., Fairclough, K., Gollan, J., Guse, N.,
Hansen, P.L., Heubeck, M., Jensen, J.-K., Le Guillou, G., Olsen, B., Olsen, K.O.,
Pedersen, J., Stienen, E.W.M., Turner, D.M., 2011. Monitoring plastic ingestion by
the northern fulmar Fulmarus glacialis in the North Sea. Environ. Pollut. 159,
2609e2615.
Van Sebille, E., England, M.H., Froyland, G., 2012. Origin, dynamics and evolution of
ocean garbage patches from observed surface drifters. Environ. Res. Lett. 7,
044040. http://dx.doi.org/10.1088/1748e9326/7/4/044040.
Vlietstra, L.S., Parga, J.A., 2002. Long-term changes in the type, but not the amount,
of ingested plastic particles in short-tailed Shearwaters in the southeastern
Bering Sea. Mar. Pollut. Bull. 44, 945e955.
Weimerskirch, H., Chastel, O., Cherel, Y., Henden, J.-A., Tveraa, T., 2001. Nest
attendance and foraging movements of northern fulmars rearing chicks at
€ rno
€ya Barents Sea. Polar Biol. 24, 83e88.
Bjo
Wong, C.S., Green, D.R., Cretney, W.J., 1974. Quantitative tar and plastic waste distribution in the Paciﬁc Ocean. Nature 247, 30e32.
Wood, S.N., 2001. mgcv: GAMs and generalized ridge regression for R. R. News 1/2,
20e25.
Wood, S.N., 2011. Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J. R. Stat. Soc. (B)
73 (1), 3e36.
Zuur, A.F., Ieno, E.N., Walker, N.J., Saveliev, A.A., Smith, G.M., 2009. Mixed Effects
Models and Extensions in Ecology with R. Statistics for Biology and Health.
Springer.
Van Franeker & Law – OnLine Supplement
page 1 of 25
ONLINE SUPPLEMENT
To:
Seabirds, gyres and global trends in plastic pollution.
By
Jan A. van Franeker & Kara Lavender Law
Author contact details:
Jan van Franeker, IMARES
PO Box 167 1790AD Den Burg-Texel, The Netherlands
E:
[email protected]
Kara Lavender Law
Sea Education Association
PO Box 6, Woods Hole
Massachusetts 02543 United States
E:
[email protected]
Contents
1. Supplementary information with introduction .......................................................................... 2 2. Supplementary information with ‘Methods Fulmar study’ ........................................................... 3 3. Supplementary information with Results retention time in fulmar stomachs ................................. 7 4. Supplementary information with Results ‘Consistency in regional patterns’ .................................. 9 5. Supplementary information on results for trends in ingested plastics ........................................ 11 6. Supplementary information on data selection for the North Atlantic subtropical gyre. .................. 15 7. Supplementary information North Atlantic subtropical gyre results ........................................... 17 8. Photographs of samples ..................................................................................................... 24 Van Franeker & Law – OnLine Supplement
page 2 of 25
1. Supplementaryinformationwithintroduction
In the early 1980’s Van Franeker (1985) observed an average of nearly 12 particles of plastic in
stomachs of Northern Fulmars Fulmarus glacialis from the North Sea, but less than 5 particles in
fulmars from the presumably cleaner, breeding locations of Bear Island (74°N-19°E) and Jan Mayen
(71°N-8°W) in polar environments of the northern Atlantic Ocean. Southern hemisphere relatives of
the fulmar breeding in the Antarctic (Ardery Island, Windmill Islands, Wilkes Land, 66°S-110°E) had
still lower levels of ingested plastics in which differences between species appeared linked to
migratory behaviour (Van Franeker & Bell 1988). Fulmarine petrels that never leave the Antarctic
seasonal sea ice zone, such as the Antarctic Petrel (Thalassoica antarctica) and Snow Petrel
(Pagodroma nivea), had zero or almost zero plastics in the stomach, whereas species known to
migrate to more populated parts of the world during winter such as the Southern Fulmar (Fulmarus
glacialoides) and especially the Cape Petrel (Daption capense) appeared to hold plastics in their
stomach during the migration.
Supplement Figure 1 Frequency of occurrence (incidence) and abundance by number of plastic
particles in stomachs of ecologically and morphologically similar species of fulmarine petrels around
the world in the 1980s. Stomach contents matched ‘expected’ differences between polluted
industrialized areas and remote areas with likely lower pollution levels.
Supplement Table 1 Abundance of plastics in stomachs of Arctic and Antarctic petrels in the
1980s (from Van Franeker 1985 and Van Franeker & Bell 1988)
n
Netherlands North Sea
Incidence % particles/bird
mg/bird
North Atlantic Fulmar populations 1982-1984
65
92%
11.9
-
Bear
Island
22
82%
4.5
-
Jan
Mayen
29
79%
4.7
-
Cape
Petrel
Southern Fulmar
Snow
Petrel
Antarctic Petrel
Wilsons Storm Petrel adults
Wilsons Storm Petrel chicks
Antarctic fulmarines - Windmill Islands 1984-1987
30
33%
1
20.1
27
7%
0.4
10.6
27
4%
+
0.1
11
0%
0
0
Wilsons Storm Petrel - Windmill Islands 1984-1987
4
75%
1.8
4.8
16
75%
5
16.4
Van Franeker & Law – OnLine Supplement
page 3 of 25
Full definition of the Ecological Quality Objective on plastic particles in seabird stomachs (OSPAR
2010b):
“There should be less than 10% of northern fulmars (Fulmarus glacialis) having
more than 0.1 g plastic particles in the stomach in samples of 50 to 100 beachwashed fulmars found from each of 4 to 5 areas of the North Sea over a period
of at least five years”.
The OSPAR EcoQO approach has not set a date for reaching such level of ‘ecological quality’. The
fulmar approach has been copied in the European Marine Strategy Directive as one of the indicators
for its descriptor 10 on Marine Litter. The equivalent wording for the EcoQO target is ‘Good
Environmental Status (GES)’ which should be achieved by 2020. Because of this date, some countries
are considering a temporary weaker target definition (e.g., significant reduction in 2020)
2. Supplementaryinformationwith‘MethodsFulmarstudy’
Fulmar monitoring for the Save the North Sea project started in 2002 and has participants that collect
beached fulmars at a number of locations around the North Sea. For comparative purposes, from the
Faroe Islands, stomachs from birds hunted, drowned in longline fisheries or collected for toxicological
work by the local Food & Environmental Agency have also been used.
Supplement Figure 2
Fulmar-Litter study sites in the
Save the North Sea Project (SNS).
Colour of symbols indicates
regional grouping into Scottish
Islands (red), East England (blue),
Channel area (white), Southeastern
North Sea (yellow), and Skagerrak
area (white). Not all locations are
equally active. The Faroe Islands
study area is considered as an
external reference monitoring site
for the North Sea.
Dissection
Bird corpses are stored frozen until analysis. At dissections, a full series of data is recorded to
determine sex, age, breeding status, likely cause of death, origin, condition index and other issues.
Age, the only variable found to influence litter quantities in stomach contents, is largely determined on
the basis of development of sexual organs (size and shape) and presence of Bursa of Fabricius (a
gland-like organ positioned near the end of the gut which is involved in immunity systems of young
birds; it is well developed in chicks, but disappears within the first year of life or shortly after).
Van Franeker & Law – OnLine Supplement
page 4 of 25
Stomach procedure
After dissection, stomachs of birds are opened for analysis. Stomachs of Fulmars have two 'units':
initially food is stored and starts to digest in a large glandular stomach (the proventriculus) after which
it passes into a small muscular stomach (the gizzard) where harder prey remains can be processed
through mechanical grinding. In early phases of the project, data for the two individual stomachs were
recorded separately, but for the purpose of reduction in monitoring costs, the contents of
proventriculus and gizzard are now combined.
Stomach, contents are carefully rinsed in a sieve with a 1 mm mesh and then transferred to a petri
dish for sorting under a binocular microscope. The 1 mm mesh is used because smaller meshes
become clogged with mucus from the stomach wall and with food-remains. Analyses using smaller
meshes were found to be extremely time consuming and particles smaller than 1 mm seemed rare in
the stomachs, and when present contribute little to plastic mass.
If oil or chemical types of pollutants are present, these may be sub-sampled and weighed before
rinsing the remainder of stomach content. If sticky substances hamper further processing of the litter
objects, hot water and detergents are used to rinse the material clean as needed for further sorting
and counting under a binocular microscope.
Categorization of debris in stomach contents
The following categorization is used for plastics and other rubbish found in the stomachs, with
acronyms between parentheses:
1. PLASTICS (PLA)
1.1.
Industrial plastic pellets (IND). These are small, often cylindrical-shaped granules of
± 4 mm diameter, but also disk and rectangular shapes occur. Various names are used, such
as pellets, beads or granules. They can be considered as “raw” plastic or a half-product in the
form of which plastics are usually first produced (mostly from mineral oil). The raw industrial
plastics are then usually transported to manufacturers that melt the granules and mix them
with a variety of additives (fillers, stabilizers, colorants, anti-oxidants, softeners, biocides, etc.)
that depend on the user product to be made. For the time being, included in data output for
this category are a relatively small number of very small, usually transparent spherical
granules also considered to be a raw industrial product.
1.2.
User plastics (USE) (all non-industrial remains of plastic objects) differentiated in the
following subcategories:
1.2.1.sheetlike user plastics (she), as in plastic bags, foils etc., usually broken up in smaller
pieces;
1.2.2.threadlike user plastics (thr) as in (remains of) ropes, nets, nylon line, packaging
straps etc. Sometimes ‘balls’ of threads and fibres form in the gizzard;
1.2.3.foamed user plastics (foa), as in foamed polystyrene cups or packaging or foamed
polyurethane in matrasses or construction foams;
1.2.4.fragments (fra) of more or less hard plastic items as used in a huge number of
applications (bottles, boxes, toys, tools, equipment housing, toothbrushes, lighters etc.);
1.2.5.other (oth), for example cigarette filters, rubber, elastics etc. Items that are ‘plastic-like’
or do not fit into a clear category.
2. RUBBISH (RUB) other than plastic:
2.1.
paper (pap) including silver paper, aluminium foil etc., so various types of non-plastic
packaging material;
2.2.
kitchenfood (kit) for human food wastes such as fried meat, chips, vegetables,
onions etc., probably mostly originating from ships’ galley refuse;
2.3.
various rubbish (rva) is used for e.g. pieces of timber (manufactured wood); paint
chips, pieces of metals etc.;
2.4.
fish hook (hoo) from either sport-fishing or long-lining.
To be able to sort out items of debris categories 1 and 2, all other materials in the stomachs have to
be sorted. Optional categories used include:
3. POLLUTANTS (POL)
Items indicating industrial or chemical waste remains such as slags (the remains of burning ovens,
e.g. remains of coal or ore after melting out the metals); tar-lumps (remains of mineral oil);
chemical (lumps or ‘mud’ of paraffin-like materials or sticky substances arbitrarily judged to be
unnatural and of chemical origin) and feather-lumps (indicating excessive preening by the bird of
feathers sticky with oil or chemical pollutants).
4. NATURAL FOOD REMAINS (FOO)
Van Franeker & Law – OnLine Supplement
page 5 of 25
Numbers of specific items may be recorded in separate subcategories (fish otoliths, eye-lenses,
squid-jaws, crustacean remains, jelly-type prey remains, scavenged tissues incl. feathers, insects,
other).
5. NATURAL NON-FOOD REMAINS (NFO)
Numbers of subcategories e.g. plant-remains, seaweed, pumice, stone and other may be
recorded.
Acronyms
In addition to the acronyms used for (sub)categories as above, further acronyms may be used to
describe datasets. Logarithmic transformed data are initiated by ‘ln’ (natural logarithm); mass data are
characterized by capital G (gram) and numerical data by N (number). For example lnGIND refers to
the dataset that uses ln-transformed data for the mass of industrial plastics in the stomachs; acronym
NUSE refers to a dataset based on the number of items of user plastics.
Particle counts and category weights
For the main categories 1 (plastic) and 2 (rubbish) we record for each bird and each (sub)category:
 The number of particles (N=count of number of items in each (sub)category)
 mass (W=weight in grams) using Sartorius electronic weighing scale after at least a two day
period of air drying at laboratory temperatures. For marine litter (categories 1 to 3 above), this
is done separately for all subcategories. In the early Fulmar study we also weighed the
natural-food and natural-non-food categories as a whole, but this was discontinued in 2006 to
reduce costs. Weights are recorded in grams accurate to the 4th decimal (= tenth of
milligram).
On the basis of these records, data can be presented in different formats:
Incidence
The simplest form of data presentation is by presence or absence. Incidence (Frequency of
occurrence) gives the percentage of investigated stomachs that contained the category of debris
discussed. The quantity of debris in a stomach is irrelevant in this respect.
Arithmetic Average
Data for numbers or mass are frequently shown as averages with standard errors (se) calculated for a
specific type of debris by location and specified time period. Averages are calculated over all available
stomachs in a sample, so including those that contained no plastic (‘population averages’). Especially
when sample sizes are smaller, arithmetic averages may be influenced by short term or local
variations or extreme outliers. An option then is to pool data over a larger area or longer time period.
An alternative to reduce the influence of outliers is by logarithmic transformation of data.
Geometric Mean
Sample sizes may not be large enough to average out the impact of occasional extreme outliers.
Therefore, data are often additionally presented as geometric means, calculated from logarithmic data
values. Logarithmic transformation reduces the role of the higher values, but as a consequence the
geometric mean is usually considerably lower than the arithmetic mean for the same data. In mass
data for plastics in the Fulmar stomachs, geometric means are only about one third to half of the
arithmetic averages. Geometric means thus do not properly reflect absolute values, but are useful for
comparative purposes between smaller sample sizes, for example when looking at annual data rather
than at 5-year-periods. Logarithmic transformation cannot deal with the value zero, and thus the
common approach chosen is to add a small value (e.g. 0.001g in mass data) to all data points, and
then subtracting this again when the mean of log values is back-calculated to normal value. However,
this implies that geometric means are less reliable with an increasing number of zero values in a
dataset. The natural logarithm (ln) is used to compute geometric means.
EcoQO performance
OSPAR (2010b) words its Ecological Quality Objective (EcoQO) for levels of litter (plastic) in stomachs
of fulmars (the ‘Fulmar-Litter-EcoQO’) as: “There should be less than 10% of northern fulmars
(Fulmarus glacialis) having more than 0.1 gram plastic particles in the stomach in samples of 50 to
100 beach-washed fulmars from each of 4 to 5 different areas of the North Sea over a period of at
least 5 years”.
Thus, the information requested for OSPAR and the EcoQO focuses on the category of ‘total plastic’
and pooled data for 5-year periods over larger areas, and a simple decision rule for each stomach if
the plastics in it weigh more than 0.1 gram or less, including zero.
EcoQO compliance or performance is defined as the percentage of birds in a sample that has 0.1 g or
more plastic mass in the stomach. The OSPAR target is thus to reduce that percentage to under 10%.
The EcoQO format is a highly simplified form of data-presentation but through that simplicity escapes
Van Franeker & Law – OnLine Supplement
page 6 of 25
the problems faced by more sophisticated procedures as a consequence of excessive outliers or a
large proportion of zero values in a data set. In the background however, details of various
subcategories of litter continue to play an important role for correct interpretation of the EcoQO metric.
Data pooling
To avoid erratic information on the level of ingested plastics from short term variations, data are
frequently pooled into 5-year periods. Such pooled data for 5-year periods are not derived from the
annual averages, but are calculated from all individual birds over the full 5-year period. For data
presentation, the Current Situation of plastic ingestion is defined as the figures for incidence and
number or mass abundance for the most recent 5-year period, not the figures for the recent single
year. Time related changes are illustrated in graphs by running 5-year averages, each time shifting
one year and thus overlapping for four years.
For pooling study locations in the North Sea, the OSPAR EcoQO target definition has triggered a
grouping into five areas or regions (Fig. 1): the Scottish Islands (Shetland and Orkney), East England
(northeast and southeast England), the Channel (Normandy and Pas de Calais), South-Eastern North
Sea (Belgium, Netherlands and Germany), and the Skagerrak (Skagen, Denmark, Lista, Norway and
Swedish west coast)
Statistical tests
Data from dissections and stomach content analysis are recorded in Excel spreadsheets and next
stored in Oracle relational database. GENSTAT 17 was used for statistical tests
(http://www.vsni.co.uk/software/genstat/). As concluded in the pilot study (Van Franeker & Meijboom
2002) and later reports, statistical trend analyses for EcoQO purposes are conducted using mass
data.
Tests for trends over time are based on linear regressions fitting ln-transformed plastic mass values
for individual birds on the year of collection. Logarithmic transformation is needed because the original
data are strongly skewed and need to be normalized for the statistical procedures. The natural
logarithm (ln) is used. Tests for ‘long term’ trends use the full data set; ‘recent’ trends only use the
past ten years of data. This 10-year period was derived from the pilot study (Van Franeker & Meijboom
2002), which found that in the Dutch situation a series of about eight years was needed to potentially
detect significant change. To be on the safe side in our approach, this period was arbitrarily increased
to a standard period of 10 years for tests of current time related trends. The test statistic is a t-score (t)
defined by t = b1 / SE where b1 is the slope of the sample regression line, and SE is the standard
error of the slope.
Statistical tests of regional differences are conducted in GENSTAT 17th edition, using mass data from
individual birds over the most recent 5-year period. Regional differences in ingested plastic mass were
evaluated by fitting a negative binominal generalized linear model with region as a factor with a log
ratio link function and estimated dispersion parameter. The test statistic is a t-score for residual
variance for the region.
Summary of Fulmar data presentation and analysis:
 Incidence – Incidence represents the percentage of birds having plastic in the stomach
 Average ± se – These refer to straightforward arithmetic averages from all available samples
(population average), usually given with standard errors.
 Geometric mean – Refers to geometric means calculated using data transformation (natural
logarithm) reducing influence of extreme outliers.
 EcoQO performance – The % of birds having more than 0.1 gram of plastic in the stomach.
 Pooled data - Data are mostly presented as pooled over 5-year periods to avoid incidental
short-term fluctuations. The ‘Current level of plastic ingestion’ is defined by pooled data for
the most recent 5 years, not by an annual figure. Graphs often use the pooled data for 5
years, but shifting one year by data point. Data points and connecting lines only intend to
visually illustrate trends over time or geographic patterns and have no statistical relevance.
 Statistics - Statistical analyses are solely based on the mass of plastic using ln-transformed
data from individual birds. Tests for significance of trends over time are based on linear
regressions of ln-transformed data against year of collection. The long-term trend is derived
from the full dataset, the Recent trend from only the most recent 10 years of data. Regional
differences were fitted in a negative binomial generalized linear model with region included as
a factor and test statistic a t-score based on residual variance for the region (Genstat 17th
Edition).
Van Franeker & Law – OnLine Supplement
page 7 of 25
3. SupplementaryinformationwithResultsretentiontimeinfulmarstomachs
Supplement Table 2
Details of plastics in the stomachs of Cape Petrels in sub-Antarctic
or temperate winter areas and an Antarctic breeding area (Windmill Islands, Wilkes Land, 66°S110°E) to illustrate disappearance rate of plastics from stomachs when feeding in clean Antarctic
waters (derived from van Franeker and Bell 1988). The single October bird compares well to winter
samples of Cape Petrels from near South Africa (Ryan 1987) and from near the Crozet Islands (Van
Franeker & J.-K. Jensen unpublished)
off S. Africa
(Ryan 1987)
at sea off
Crozet
Islands 1
Aug 1998
off Windmill
Islands 23
Oct 1986
18
13
1
32
9
20
plastic incidence
83%
100%
100%
91%
56%
20%
38%
64%
average number of items per bird
8.60
11.46
11.00
9.84
1.67
0.25
83%
85%
average mass per bird (g)
0.106
0.045
0.294
0.15
0.027
0.003
82%
88%
CAPE PETREL
sample size
pre-breed
Windmill
Windmill pre-breed to December
combined Islands 1-10 Islands 20- December to January
decrease % decrease %
estimate
Dec-1986
Jan-1986
Note: In the 1986/87field season on Ardery Island, among bird bones around Skua nests (a species predating on
fulmarine petrels) two remains of Cape Petrels were found with bands from New Zealand indicating wintering in
more temperate waters with higher plastic pollution levels.
Supplement Table 3
Decrease in squid beak abundance in stomachs of fulmarine petrels
in the Windmill Islands, Antarctica 1985-1987 from early incubation period to middle of the
breeding season; derived from Van Franeker et al., 2001*, Table 7). ). Snow Petrels had low
abundance of squid beaks to start with, because they winter in the sea ice close to the breeding
areas.
December
n
avg nr of
squid
beaks in
stomach
Southern Fulmar
6
Antarctic Petrel
January
n
avg nr of
squid
beaks in
stomach
%
decrease
9.5
21
3.6
62%
5
7.6
6
1.2
84%
Cape Petrel
9
11.1
20
1.1
90%
Snow Petrel (major)
4
1.5
13
0.9
40%
Snow Petrel (nivea)
7
1.6
2
1
38%
fulmarine petrels combined
31
6.8
62
1.9
72%
(* Van Franeker, J.A., Williams, R., Imber, M.J., Wolff, W.J., 2001. Diet and foraging ecology of Southern Fulmar
Fulmarus glacialoides, Antarctic Petrel Thalassoica antarctica, Cape Petrel Daption capense and Snow
Petrels Pagodroma nivea ssp on Ardery Island, Wilkes Land, Antarctica.. Chapter 11 (58pp) in: J.A. van
Franeker Mirrors in ice. PhD-Thesis, University of Groningen. Alterra, Texel. ISBN 90-367-1352-8.)
Notes on rates of disappearance of plastic debris from stomachs in other species
The conclusion for fulmars and allies that an estimated 75% of harder plastics in a stomach are
ground within a month to dust and sizes small enough to pass in to the intestines cannot be
generalized, as this will depend on the size, stomach morphology and feeding habits of the species
Van Franeker & Law – OnLine Supplement
page 8 of 25
studied. For Northern Fulmars, preliminary data (Bravo-Rebolledo 2011) indicate that particles need to
wear down to approximately 2 mm in diameter before passage from gizzard to intestine becomes
possible. Once small enough for such passage, the remaining bits can make their way through the
intestines in a matter of hours and are excreted back into the environment, only in a smaller size range
than originally ingested. The speed of intestinal passage means that hard items including plastics are
only incidentally found in the intestines of fulmarine petrels.
This situation can be different even in related and morphologically similar, but differently sized
seabirds. For example, chicks of Wilsons Storm-petrels (Oceanites oceanicus) found in the Windmill
Island area in Antarctica during the mid-1980s had considerable quantities of plastic in the stomach,
mainly the gizzard (Van Franeker & Bell 1988; Online Supplement Table 1). The bulk of these plastics
came from ingestion of plastics by the parent birds in wintering areas at least ~3 months earlier, and
had no relation to local pollution levels. However, to be fed to chicks, plastics must have been residing
in the parental proventriculus, where no grinding occurs, over the egg incubation period, and once in
gizzards of chicks, the small size of these birds may require that plastics have to be ground to much
smaller sizes before they can pass into the intestines. Industrial granules have their original size
mostly in the range of 4 to 5 mm. Fulmars likely can excrete the granules earlier at relatively larger
residual size than storm petrels. Average mass of remaining pellets in stomachs of fulmars from the
North Sea is around 0.023 gram, whereas pellets remaining in Wilsons Storm-petrels in the Antarctic
were extremely worn and only weighed ~0.005 g on average
Van Franeker & Law – OnLine Supplement
page 9 of 25
4. SupplementaryinformationwithResults‘Consistencyinregionalpatterns’
Supplement Figure 3
Map of the north Atlantic showing locations mentioned in relation to
spatial patterns of plastic ingestion by northern fulmars (Supplement Table 4 and Supplement
Table 5).
Supplement Table 4
Large-scale latitudinal patterns in plastic ingestion in Northern
Fulmars from different areas in the North Atlantic and Pacific Oceans. Data from a) Bond et al.
2014, b) this study, c) Kühn & van Franeker 2012, d) combined from Mallory et al 2006, Mallory
2008, Provencher et al 2009 and personal information, e) Nevins et al. 2011, and f) Avery-Gomm
et al. 2012)
North Sea
2007-2011
Faroe Islands
2007-2011
Iceland
2011
Atl. Arctic
Canada 20022008
816
95%
699
91%
58
79%
169
40%
11.3
0.6
0.15
0.01
6.0
1.0
0.13
0.04
2.5
se
32.8
2.2
0.37
0.03
EcoQO performance *
62%
40%
28%
n
incidence
nr of plastic particles per stomach
se
mass of plastic per stomach (g)
* EcoQO performance = percentage of birds having more than 0.1g of plastic in the stomach
California
94%
67
93%
Alaska
63%
36.8
9.8
0.39
0.09
0.03
14%
Eastern North
Pacific
89%
61%
25%
Van Franeker & Law – OnLine Supplement
page 10 of 25
Supplement Table 5
Regional EcoQO performance (proportion of Fulmars having > 0.1 g
plastic in the stomach) in sub-regions within the North Sea, and on Faroe Islands, in running 5year periods 2002 to 2011, as displayed in Figure 3 of the main paper.
76%
71%
65%
64%
63%
63%
60%
58%
57%
58%
59%
60%
59%
61%
59%
56%
115
222
258
304
329
294
318
331
304
309
290
227
212
204
223
80%
67%
0%
41
1990-1994
60%
50%
1993-1997
1994-1998
1995-1999
1996-2000
1997-2001
1998-2002
1999-2003
2000-2004
2001-2005
2002-2006
2003-2007
2004-2008
2005-2009
2006-2010
2007-2011
2008-2012
1
1989-1993
3
5
1988-1992
87%
79%
80%
78%
77%
72%
74%
71%
71%
10
6
1979-1983
1980-1984
1981-1985
1982-1986
1983-1987
1984-1988
1985-1989
1986-1990
1987-1991
1992-1996
23
42
45
49
61
43
27
24
21
5-year period
Inc.
%
1991-1995
sample
n
NETHERLANDS
± 1.8
± 1.6
± 1.5
± 1.4
± 1.2
± 1.5
± 1.2
± 1.4
± 1.5
3.8
3.6
3.6
3.5
3.4
3.3
3.0
2.8
2.9
2.6
2.8
2.6
4.1
4.4
3.8
± 0.6
± 0.5
± 0.5
± 0.4
± 0.4
± 0.4
± 0.3
± 0.3
± 0.3
± 0.2
± 0.3
± 0.3
± 1.3
± 1.4
± 1.3
5.1 ± 1.5
2.6 ± 0.9
1.0 ± 0.6
0.0 ± 0.0
4.2 ± 2.5
3.5 ± 2.1
8.0
8.9
8.6
8.0
7.1
6.2
3.6
3.4
3.8
avg number
n ± se
± 0.04
± 0.03
± 0.03
± 0.03
± 0.03
± 0.03
± 0.03
± 0.04
± 0.04
0.09
0.07
0.08
0.07
0.07
0.07
0.06
0.06
0.06
0.06
0.06
0.06
0.09
0.10
0.08
± 0.01
± 0.01
± 0.01
± 0.01
± 0.01
± 0.01
± 0.01
± 0.01
± 0.01
± 0.01
± 0.01
± 0.01
± 0.03
± 0.03
± 0.03
0.12 ± 0.03
0.06 ± 0.02
0.02 ± 0.01
0.00 ± 0.00
0.11 ± 0.06
0.09 ± 0.06
0.17
0.18
0.18
0.17
0.15
0.14
0.10
0.09
0.10
avg mass
g ± se
Industrial granules
96%
96%
97%
97%
97%
97%
94%
94%
93%
92%
92%
94%
93%
94%
93%
98%
100%
100%
100%
100%
100%
87%
88%
89%
88%
84%
81%
78%
75%
76%
Inc.
%
± 1.5
± 1.7
± 1.6
± 1.4
± 1.3
± 1.7
± 1.8
± 2.0
± 2.3
26.5
29.1
27.7
26.5
29.7
30.2
25.8
25.1
26.4
23.8
24.7
24.8
30.5
31.1
26.0
± 3.9
± 3.4
± 3.0
± 2.6
± 3.1
± 3.4
± 2.7
± 2.6
± 2.8
± 2.0
± 2.1
± 2.3
± 3.2
± 3.4
± 2.9
27.5 ± 5.8
20.3 ± 11.1
6.0 ± 2.5
11.0 ± 0.0
11.0 ± 5.0
9.5 ± 4.4
6.9
7.7
7.5
7.3
7.7
7.8
7.9
8.3
9.0
± 0.10
± 0.07
± 0.07
± 0.06
± 0.05
± 0.04
± 0.03
± 0.03
± 0.04
0.74
0.57
0.53
0.47
0.45
0.30
0.24
0.23
0.24
0.22
0.24
0.23
0.23
0.24
0.23
± 0.23
± 0.13
± 0.11
± 0.10
± 0.09
± 0.06
± 0.04
± 0.04
± 0.04
± 0.02
± 0.02
± 0.02
± 0.02
± 0.02
± 0.04
0.49 ± 0.13
0.16 ± 0.08
0.07 ± 0.04
0.14 ± 0.00
0.15 ± 0.09
0.14 ± 0.07
0.32
0.25
0.25
0.23
0.18
0.13
0.10
0.10
0.10
avg mass
g ± se
User plastics
avg number
n ± se
97%
97%
98%
97%
98%
98%
95%
95%
94%
93%
93%
95%
94%
95%
94%
98%
100%
100%
100%
100%
100%
100%
95%
96%
94%
90%
86%
85%
83%
86%
± 2.2
± 2.9
± 2.7
± 2.5
± 2.2
± 3.0
± 2.6
± 3.0
± 3.3
30.3
32.7
31.3
29.9
33.1
33.5
28.8
27.9
29.3
26.5
27.4
27.3
34.5
35.5
29.7
± 4.0
± 3.7
± 3.2
± 2.8
± 3.3
± 3.6
± 2.9
± 2.7
± 3.0
± 2.1
± 2.2
± 2.5
± 3.8
± 4.0
± 3.6
32.6 ± 6.1
22.9 ± 11.2
7.0 ± 2.1
11.0 ± 0.0
15.2 ± 5.2
13.0 ± 4.8
14.9
16.5
16.1
15.3
14.9
14.0
11.5
11.7
12.8
± 0.11
± 0.09
± 0.08
± 0.07
± 0.06
± 0.07
± 0.06
± 0.07
± 0.07
0.83
0.64
0.60
0.55
0.52
0.37
0.30
0.29
0.30
0.28
0.30
0.29
0.32
0.33
0.31
± 0.23
± 0.13
± 0.12
± 0.10
± 0.10
± 0.06
± 0.04
± 0.04
± 0.04
± 0.02
± 0.03
± 0.03
± 0.04
± 0.04
± 0.05
0.61 ± 0.13
0.22 ± 0.09
0.08 ± 0.03
0.14 ± 0.00
0.26 ± 0.14
0.23 ± 0.12
0.50
0.43
0.42
0.40
0.34
0.26
0.20
0.19
0.21
average mass
g ± se
Total plastics
Incidence average number
n ± se
%
0.184
0.151
0.149
0.137
0.130
0.112
0.095
0.091
0.094
0.092
0.096
0.102
0.107
0.110
0.091
0.217
0.081
0.072
0.140
0.150
0.122
0.298
0.154
0.159
0.137
0.100
0.062
0.063
0.052
0.063
Geometric
mean mass
73%
67%
67%
63%
62%
59%
59%
57%
61%
61%
62%
58%
62%
60%
54%
76%
50%
33%
100%
80%
67%
91%
74%
76%
71%
66%
53%
56%
50%
57%
EcoQO %
(over 0.1g)
Van Franeker & Law – OnLine Supplement
page 11 of 25
5. Supplementaryinformationonresultsfortrendsiningestedplastics
Supplement Table 6
Running 5-year averages of plastic abundance in Fulmars from the
Netherlands since the 1980s. Data show incidence, and arithmetic averages ± se for number of
particles and mass for each of the major categories and total of plastics in running 5-year averages
(i.e. data points shift one year ahead at a time). For all plastics combined, geometric mean mass
and EcoQO performance are added. Five-year periods shaded grey should not be used because of
small sample sizes (n<20).
Van Franeker & Law – OnLine Supplement
page 12 of 25
Supplement Figure 4
Geometric mean of mass of industrial and user plastics in Fulmars
from the Netherlands since the 1980s, illustrating the impact that strong outliers can have on even
5-year arithmetic averages: (i.e., the high values for industrial plastics in recent years in Fig.5 in
the main paper are not seen here in the geometric means).
SURVEY BY GAMM OF TIME TRENDS IN PLASTICS IN FULMAR STOMACHS
Because of its policy perspective, the Fulmar EcoQO approach (Van Franeker et al., 2011) has its main
focus on the direction and significance of recent trends (past 10 years) as reflected by linear regression
analyses of mass of plastic. However, for the long-term changes in particle number and mass considered
in this paper, data patterns in long-term trends are the main focus, and these were evaluated by
Generalized Additive Mixed Models (GAMM) using R software Version 3.0.3. Where GAMM indicates
estimated degrees of freedom (edf) at or close to value 1, the trend can be considered as a linear
relationship. Higher edf values indicate more complicated non-linear trends. Data in Supplement Table 7
show that changes over the time-period 1979-2012 for number and mass of industrial plastics are best
considered as linear correlations, but that user plastics have undergone more complicated non-linear
changes, which however differ for number of user particles and user plastic mass (Supplement Figure 5
A, B). For all plastics combined, trends for numerical abundance (Supplement Figure 5 C) are similar to
those of user plastics, but a linear relation is indicated for mass.
The GAMM patterns match those revealed by the running 5-year averages for user plastics in fulmar
stomachs (Figure 5 in the paper): the number of particles increased strongly in early years, and
remained fairly constant after the 1990s. However, mass data do show a similar initial increase, but then
decreased as a consequence of reduced sizes of user plastics in fulmar stomachs.
Supplement Table 7 Output from GAMM analyses for trends in plastic abundance by number or
mass in fulmar stomachs 1979-2012 (n=973). Shown are results for only term ‘year’. Additional
smooth terms such as age and sex were also tested but produced no different conclusions.
Bracketed letters refer to graphs showing non-linear trends in Supplement Figure 6.
GAMM analysis ~year
SIGNIFICANCE OF SMOOTH TERM YEAR
F
p
edf
1
1
3.466
0.063
2.052
0.152
Number of user particles
Mass of user particles
2.3
2.9
5.006
0.005
(A)
3.977
0.009
(B)
Total nr of plastic particles
Total mass of plastics
1.6
1
3.071
0.058
(C)
3.358
0.067
Number of industrial pellets
Mass of industrial pellets
Van Franeker & Law – OnLine Supplement
page 13 of 25
Supplement Figure 5
Non-linear trends over time for plastics in fulmar stomachs in the
Netherlands North Sea 1979-2012, as suggested by GAMM for A. number of user particles, B. user
plastic mass, and C. total number of plastic particles
Supplement Table 8
Linear regression results for plastic abundance in stomachs of
fulmars in the Netherlands 1979-2012. Note that GAMM analyses indicate that only industrial
plastics (by number and mass) and total plastics (by mass) follow linear trend patterns, whereas
the others show variable non-linear patterns (plus or minus signs indicate increasing resp.
decreasing linear trends and significance levels as ns=not significant and number of symbols
reflecting significance levels at p<0.05, p<0.01 and p<0.001)
A.
LONG TERM TRENDS 1979-2012 FOR NUMBER OF PLASTIC PARTICLES
in Fulmar stomachs, the Netherlands
n
Industrial plastics (lnNIND)
973
User plastics (lnNUSE)
973
All plastics combined (lnNPLA)
973
Constant estimate
48.5
-0.0238
-44.8
0.0236
-14.7
0.0086
s.e.
0.0045
0.0061
0.0061
t
p
-5.29 <0.001
3.90 <0.001
1.43 0.154
--+++
ns
s.e.
0.0105
0.0090
0.0088
t
-4.63
0.63
-2.07
--ns
-
B.
LONG TERM TRENDS 1979-2012 FOR PLASTIC MASS
in Fulmar stomachs, the Netherlands
n
Industrial plastics (lnGIND)
973
User plastics (lnGUSE)
973
All plastics combined (lnGPLA)
973
Constant estimate
93.1
-0.0486
-14.0
0.0057
34.1
-0.0182
p
<.001
0.531
0.039
Data analyses are complicated by the fact that user plastic particles have shown a change in particle
size over time. This is evident from Fulmar data (see Supplement Figure 6), but was also reported for
the North Atlantic subtropical gyre by Moret-Ferguson et al. (2010).
Van Franeker & Law – OnLine Supplement
page 14 of 25
Supplement Figure 6
Changes over time in user plastic particle size (mass), but not in
industrial pellets in stomachs of fulmars from the Netherlands. Data derived from annual sums of
mass divided by number of particles. We are currently studying the size and mass of individual
particles, but these data are not yet available.
Van Franeker & Law – OnLine Supplement
page 15 of 25
6. SupplementaryinformationondataselectionfortheNorthAtlantic
subtropicalgyre.
Supplement Figure 7
SEA neuston net trawl station positions 1986-2012. One-time
transects east of 50°West were ignored. Data analysis focused on the ‘central gyre’ arbitrarily
defined as area 20-40°N and 60-80°W, which avoids the shelf and Caribbean areas and which has
average plastic debris densities about three times the level seen in the other, ‘peripheral’ areas.
Supplement Figure 8
Data survey to explore impact of large number of zero observations
and extreme outliers on average densities calculated as either ‘sum of particles / sum of area
surveyed’ per year (circles, no standard errors) or as averages calculated from individual tow
densities during the year (triangles ± standard errors). (Data for all of the Atlantic Gyre west of
Van Franeker & Law – OnLine Supplement
page 16 of 25
50°W, and no exclusion of likely erroneous sample counts with industrial >=10 but user=0). We
conclude that both modes of calculation produce very similar averages, and here the individual tow
calculation with standard errors for the means is preferred.
Supplement Figure 9
Background of exclusion of tow data when more than 10 industrial
granules were reported but zero user plastic particles. The percentage of industrial granules
among all plastic particles in the gyre averages below 7%. Assuming a 10% occurrence, the
likelihood to obtain a tow with 4 industrial particles but zero user plastics approaches zero (red
dashed line). We excluded tows that reported 10 or more industrial granules but zero user plastics
as such combinations are utterly unlikely and suspected of being data entry errors. The line for
observed frequencies is based on all counts west of 50°W; 6844 tows; 5429 industrial granules and
78697 user particles in nets).
Van Franeker & Law – OnLine Supplement
page 17 of 25
7. SupplementaryinformationNorthAtlanticsubtropicalgyreresults
Supplement Table 9
5-year running averaged data and linear regression results for the
central part of the North Atlantic subtropical gyre, excluding tows with more than 10 industrial
granules but zero user plastics. This dataset was mostly used in the paper to illustrate trends. See
next pages for comparative information on full data set and annual figures for central part and
periphery of the gyre
Van Franeker & Law – OnLine Supplement
page 18 of 25
Supplement Table 10
Output from GAMM analyses for trends in plastic abundance by
number in the central part of the North Atlantic subtropical gyre, n=2624 tows 1987-2012.
Additional smooth terms such as latitude and longitude were also tested but produced no different
conclusions. Where GAMM indicates estimated degrees of freedom (edf) at or close to value 1, the
trend can be considered as a linear relationship (A) and (B) refer to plotted trends in Supplement
Figure 10.
SIGNIFICANCE OF SMOOTH TERM YEAR
GAMM analysis ~year
Number of industrial pellets
Number of user particles
Total nr of plastic partices
edf
F
p
1
35.97
<0.001
6.1
6.383
<0.001
(A)
6.2
5.849
<0.001
(B)
Supplement Figure 10
Complex non-linear trends over time as suggested by
GAMM for user plastic (A) and total plastic (B) abundance in the central part of the North Atlantic
subtropical gyre. Extreme outliers in individual years render any analysis complicated.
Supplement Table 11
Linear regression results for numerical plastic abundance in central
part of the North Atlantic subtropical gyre. GAMM analysis indicates indeed linear fit for industrial
pellets, but a complicated non-linear pattern for user particles, which in linear analysis shows no
noticeable overall change (plus or minus signs indicate increasing resp. decreasing linear trends
and significance levels as ns=not significant and number of symbols reflecting significance levels at
p<0.05, p<0.01 and p<0.001)
Linear regression on densities of plastics in the central N.Atlantic Subtropical Gyre 1987-2012
n
Constant
estimate
s.e.
t
p
Industrial plastics
2624
258.5
-0.1283
0.0098
-13.05
<0.001
---
User plastics
2624
24.9
-0.0093
0.0122
-0.76
0.448
ns
All plastics combined
2624
30.7
-0.0121
0.0119
-1.01
0.313
ns
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
YEAR
52
0
19
62
104
199
156
102
161
162
165
144
172
162
178
94
59
60
74
46
87
59
59
78
69
101
Nr of
tows
65%
47%
48%
43%
42%
40%
32%
34%
39%
33%
22%
24%
22%
23%
21%
15%
10%
18%
9%
20%
22%
15%
10%
15%
12%
0.03
0.11
0.17
0.36
0.27
0.19
0.27
0.26
0.27
0.26
0.28
0.29
0.32
0.16
0.09
0.09
0.14
0.08
0.16
0.10
0.12
0.16
0.14
0.20
Incidence
(% of tows)
0.09
Surface
sampled
(k m2)
2861
705
1226
723
1095
793
803
1239
774
326
436
668
729
538
1011
320
305
125
373
486
118
70
417
165
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
1400
143
246
94
250
240
147
257
156
79
80
406
233
167
485
165
128
75
128
184
55
26
190
55
1826 ± 474
n/km2 ± se
41
29
24
19
18
9
11
18
10
3
5
3
4
4
2
1
2
1
3
4
2
1
2
1
126
geom.
mean
n/km2
INDUSTRIAL PELLETS
74%
84%
72%
71%
66%
59%
76%
82%
79%
72%
78%
78%
73%
68%
56%
65%
73%
72%
71%
46%
78%
76%
71%
89%
96%
Incidence
(% of tows)
12653
5231
7748
5592
9227
4319
11402
16883
14085
5885
11934
19870
12715
18467
14716
6112
10899
4552
12014
3369
6550
11263
4448
29195
± 6390
± 1060
± 1264
± 696
± 1438
± 790
± 2103
± 2469
± 3715
± 927
± 1590
± 5217
± 1897
± 7225
± 6147
± 1715
± 2337
± 1610
± 4021
± 1514
± 1263
± 2203
± 1587
± 6679
11948 ± 3686
n/km2 ± se
USER PARTICLES
525
874
521
379
270
134
749
1747
940
353
943
864
635
377
121
215
623
275
447
32
656
661
252
2840
3445
geom.
mean
n/km2
74%
86%
73%
74%
70%
63%
80%
83%
81%
74%
80%
79%
75%
75%
56%
70%
74%
74%
77%
54%
78%
76%
75%
89%
96%
Incidence
(% of tows)
15514
5936
8974
6315
10322
5112
12204
18122
14860
6211
12369
20537
13444
19005
15727
6432
11204
4677
12387
3854
6667
11333
4864
29360
± 7687
± 1158
± 1426
± 751
± 1511
± 857
± 2129
± 2525
± 3820
± 956
± 1607
± 5230
± 1948
± 7240
± 6185
± 1705
± 2349
± 1628
± 4014
± 1632
± 1262
± 2210
± 1613
± 6687
13774 ± 4128
n/km 2 ± se
TOTAL PLASTICS
604
1131
616
508
403
196
1089
2058
1231
426
1167
983
768
643
139
335
716
325
724
65
679
664
365
2926
4038
geom.
mean
n/km2
20%
12%
13%
12%
11%
15%
7%
7%
5%
5%
4%
3%
4%
2%
6%
4%
2%
3%
3%
13%
2%
1%
9%
1%
14%
Proportion
Industrial
Central Atlantic Gyre - annual averages (area 20°N-40°N and 60°W-80°W, unlikely entries with >10 industrial but 0 user plastics excluded)
Van Franeker & Law – OnLine Supplement
page 19 of 25
Supplement Table 12
Annual averaged data for A) Central gyre, B) peripheral areas
around the gyre, and C) ALL tows in the North Atlantic gyre west of 50°W including unlikely tows
with more than 10 industrial granules but zero user plastics.
A)
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Period
Nr of
tows
0
0
60
0
72
8
188
296
206
208
183
245
218
270
175
274
193
151
141
133
162
124
166
102
89
209
193
127
13%
13%
25%
21%
19%
13%
13%
10%
14%
12%
8%
12%
13%
13%
9%
10%
7%
16%
10%
12%
13%
6%
13%
8%
4%
0.14
0.01
0.34
0.52
0.37
0.38
0.34
0.38
0.36
0.47
0.28
0.50
0.33
0.26
0.24
0.21
0.28
0.22
0.31
0.19
0.18
0.42
0.38
0.23
Incidence
(% of tows)
428
607
444
401
218
154
147
217
224
129
261
237
223
260
187
110
557
80
188
140
64
208
76
49
± 303
± 472
± 105
± 74
± 54
± 34
± 56
± 53
± 62
± 40
± 94
± 55
± 62
± 156
± 61
± 44
± 303
± 24
± 58
± 46
± 38
± 48
± 23
± 25
119 ± 52
n/km2 ± se
-
1
6
3
3
2
1
1
2
1
1
1
1
1
1
1
1
2
1
1
1
0
2
1
0
1
geom.
mean
n/km2
INDUSTRIAL PELLETS
(outside 0°N- 40°N and 60°W - 80°W)
0.12
Surface
sampled
(k m2)
periphery of the Atlantic Gyre
78%
63%
53%
56%
35%
42%
55%
54%
46%
41%
57%
52%
58%
40%
55%
32%
66%
57%
49%
43%
57%
71%
74%
68%
50%
Incidence
(% of tows)
3688
1282
2036
2882
981
1097
2100
4467
3501
2323
2635
2561
3785
2389
3220
783
8142
2289
3198
661
4828
20240
3949
15119
± 760
± 637
± 324
± 381
± 203
± 171
± 320
± 921
± 810
± 519
± 425
± 375
± 878
± 534
± 879
± 153
± 2497
± 393
± 908
± 126
± 1967
± 2943
± 571
######
1789 ± 726
n/km2 ± se
USER PARTICLES
411
91
54
76
12
21
68
72
38
22
78
54
91
22
66
9
213
75
40
19
84
485
317
212
38
geom.
mean
n/km2
78%
63%
58%
59%
37%
46%
57%
57%
47%
43%
58%
57%
61%
44%
57%
37%
70%
58%
55%
49%
57%
71%
76%
68%
53%
Incidence
(% of tows)
4116
1890
2480
3283
1199
1251
2247
4684
3724
2452
2897
2799
4008
2649
3407
893
8699
2369
3387
801
4892
20449
4025
15169
± 881
± 1090
± 398
± 420
± 229
± 186
± 343
± 944
± 833
± 543
± 453
± 394
± 910
± 577
± 887
± 156
± 2511
± 399
± 907
± 138
± 1975
± 2964
± 574
± 10515
1908 ± 725
n/km 2 ± se
TOTAL PLASTICS
Annual Averages unlikely entries w ith >10 industrial but 0 user plastics excluded)
436
101
80
101
15
28
81
94
42
26
91
80
117
29
81
14
296
85
65
31
85
509
362
213
51
geom.
mean
n/km2
9%
32%
19%
13%
19%
13%
7%
5%
7%
5%
9%
8%
6%
13%
6%
11%
7%
3%
6%
19%
1%
1%
2%
0%
5%
Proportion
Industrial
Van Franeker & Law – OnLine Supplement
page 20 of 25
B)
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
YEAR
Nr of
tows
0
0
112
0
91
70
292
497
362
313
348
407
384
415
350
437
371
245
200
197
243
170
253
162
148
287
262
228
38%
20%
46%
29%
28%
25%
20%
22%
24%
22%
13%
19%
16%
18%
14%
12%
10%
19%
9%
15%
17%
10%
13%
10%
8%
0.17
0.13
0.51
0.88
0.64
0.57
0.62
0.64
0.64
0.74
0.57
0.79
0.65
0.43
0.33
0.31
0.42
0.31
0.47
0.29
0.30
0.58
0.52
0.43
Incidence
(% of tows)
0.22
Surface
sampled
(k m2)
936
693
723
557
596
680
650
624
475
222
630
421
465
367
430
377
1438
92
252
306
85
171
166
101
± 387
± 136
± 113
± 61
± 114
± 202
± 138
± 110
± 78
± 45
± 243
± 156
± 117
± 115
± 151
± 136
± 557
± 27
± 58
± 84
± 32
± 35
± 54
± 28
911 ± 235
n/km2 ± se
3
24
7
7
5
3
4
5
4
1
3
2
3
2
1
1
3
1
2
2
1
1
1
1
14
geom.
mean
n/km2
INDUSTRIAL PELLETS
77%
81%
60%
62%
48%
47%
64%
65%
60%
52%
67%
62%
65%
51%
55%
41%
66%
61%
57%
44%
66%
72%
73%
77%
71%
Incidence
(% of tows)
5560
4780
4070
3956
4534
2137
6379
9409
8040
3553
7182
8972
8070
8558
6611
2390
8747
2902
6230
1643
5514
17801
4080
21354
± 1488
± 953
± 520
± 364
± 665
± 294
± 1017
± 1166
± 1680
± 473
± 846
± 1985
± 1043
± 2827
± 1942
± 558
± 1812
± 524
± 1524
± 564
± 1284
± 2236
± 591
± 6561
6506 ± 1811
n/km2 ± se
USER PARTICLES
433
675
122
143
46
37
198
258
150
57
257
151
231
65
79
24
253
107
92
23
191
527
298
669
312
geom.
mean
n/km2
77%
83%
63%
65%
51%
52%
68%
67%
62%
54%
69%
65%
68%
56%
57%
48%
72%
62%
63%
51%
66%
73%
76%
77%
73%
Incidence
(% of tows)
6496
5474
4793
4512
5130
2817
7029
10033
8515
3775
7813
9393
8535
8925
7042
2767
10185
2994
6482
1949
5600
17971
4246
21455
± 1786
± 1043
± 596
± 396
± 704
± 364
± 1038
± 1198
± 1728
± 492
± 884
± 1993
± 1074
± 2837
± 1959
± 567
± 1875
± 531
± 1522
± 609
± 1289
± 2251
± 598
± 6567
7417 ± 2025
n/km2 ± se
TOTAL PLASTICS
Atlantic Gyre - Annual data for all tows west of 50°W (including unlikely high industrial -zero user counts)
466
859
166
197
64
57
287
322
182
70
335
206
289
96
95
44
437
122
149
42
196
547
363
680
392
geom.
mean
n/km2
14%
13%
15%
13%
12%
24%
10%
6%
6%
6%
8%
4%
5%
4%
6%
16%
13%
3%
4%
18%
1%
1%
4%
0%
12%
Proportion
Industrial
Van Franeker & Law – OnLine Supplement
page 21 of 25
C)
Van Franeker & Law – OnLine Supplement
page 22 of 25
A. Centre
B. Periphery
C. Total
Supplement Figure 11
Annual figures for incidence and densities of industrial and user
plastic in the North Atlantic subtropical gyre. A. central part of gyre b. peripheral parts of gyre C.
total (cf. Supplement Table 12 )
Van Franeker & Law – OnLine Supplement
page 23 of 25
Supplement Table 13
Linear tests for trends in annual densities of industrial and user
plastic in the North Atlantic subtropical gyre. A. central part of gyre b. peripheral parts of gyre C.
total (cf. Supplement Table 12). Industrial plastics decreasing in all three datasets. Increase in
user plastics suggested from data in periphery probably linked to few outliers in recent years, and
not present in the overall data set combining central and peripheral data.
A.
Linear regression on annual densities of plastics in the central N.Atlantic Gyre (1987-2012; 2624 hauls)
n
Constant estimate
s.e.
t
p
121385
-60.30
11.60
-5.22
<0.001
Industrial plastics
25
User plastics
25
-216304
114.00
167.00
0.68
0.503
All plastics combined
25
-94919
53.00
170.00
0.31
0.758
--ns
ns
B).
Linear regression on annual densities of plastics in the periphery of the North Atlantic Gyre (1987-2012; 4193 hauls)
n
Constant estimate
s.e.
t
p
18711
-9.24
3.47
-2.66
0.014
Industrial plastics
25
User plastics
25
++
-612586
308.00
108.00
2.86
0.009
All plastics combined
25
+
-593876
299.00
109.00
2.75
0.011
C.
Linear regressions on annual densities of plastics in the North Atlantic Gyre (1987-2012; 6844 hauls)
n
Constant estimate
s.e.
t
p
52677
-26.09
6.93
-3.77
0.001
Industrial plastics
25
User plastics
25
-436546
222.00
117.00
1.89
0.071
All plastics combined
25
-383870
196.00
118.00
1.65
0.112
--ns
ns
Van Franeker & Law – OnLine Supplement
page 24 of 25
8. Photographsofsamples
Photo examples of Fulmar stomach contents
Sample NMD-2010-001 with plastics from proventricular stomach contents on left, and gizzard
contents on the right. Preproduction plastic pellets for each subsample on top right are 4 to 5 mm in
diameter.
Sample FAE-2011-X29, with single preproduction plastic pellet (diameter 4-5mm) at top left.
Van Franeker & Law – OnLine Supplement
page 25 of 25
Photo examples of SEA neuston net plastic particles
photo: SEA - G. Proskurowski
photo
SEA - J. Donahue